Zheng Zhang, Huazhong University of Science and Technology, China
Xinyu Shao, Huazhong University of Science and Technology, China
Daoyuan Yu, Huazhong University of Science and Technology, China
Peigen Li, Huazhong University of Science and Technology, China
This paper studies the problem of attribute discernibility reduction and introduces some new concepts to Rough Set theory, such as Secondary Core, Regeneration Matrix and the degree of attribute discernibility. An attribute reduction algorithm has been proposed based on the concept of highest discernibility. The efficacy of this algorithm has been verified by practical application in a diagnostic system of mechanical loader, which displays notably decrease in information gathering requirement and lowers down the overall cost without any loss of accuracy.
Citation:
Zheng Zhang, Xinyu Shao, Daoyuan Yu, Peigen Li, "A Reduction Algorithm Based on the Degree of Attribute Discernibility," isda, vol. 1, pp.1009-1015, Sixth International Conference on Intelligent Systems Design and Applications (ISDA'06) Volume 1, 2006